Non-insulin dependent diabetes mellitus (NIDDM) has reached an epidemic proportion in the United States. Although NIDDM, cardiovascular diseases, and cancer account for two-thirds of all deaths in the United States, there is strong evidence to indicate that these diseases may be related to the lifestyle of the patients. There is no compelling evidence in the literature that the following are combined, or independent risk factors for development of NIDDM: (a) obesity, (b) family history of NIDDM, (c) upper body adiposity, (d) ethnicity, (e) hyperinsulinemia, (f) impaired glucose tolerance, (g) gestational diabetes, (h) sex hormone binding globulin (SHBG), (i) sedentary life. We hypothesize that in such individuals with high risk, alteration of lifestyle, such as dietary modification and physical exercise, will ameliorate or delay development of NIDDM. We, therefore, propose the following specific aims for this multicenter primary prevention trial: 1. To recruit a cohort of subjects at high risk for NIDDM consisting of 100 persons with previous history of gestational diabetes, most of whom will be African American, and 100 other persons who are hyperinsulinemic with upper body adiposity, insulin resistant, impaired glucose tolerant and strong family history of diabetes. Some of the patients will be undiagnosed NIDDM with fasting blood glucose of < 140 mg/dl. 2. To randomize these subjects into intensive therapy group versus usual care group (attention control). 3. The intensive therapy group will be designed to accomplish the following aims: (a) to modify the diets in these high risk subjects to reduce total fat to less than 30% of total calories and saturated fat to less than 10%, (b) to increase energy expenditure from physical activity to 2000 Kcal per week, (c) to combine dietary therapy with effective moderate exercise therapy to achieve a reduction of body weight of greater than 10% per individual which will be maintained over time, (d) to design these dietary and exercise interventions so they are flexible enough that they can be modified for the different target ethnic, gender, educational level, and other subgroups, and (e) to design a long term adherence program that will maximize adherence to prescribed therapies while minimizing drop outs and therapeutic cross overs. 4. To provide baseline and semi-annual evaluations of glycemic control and insulin resistance in all groups of patients, and repetition of all initial laboratory and physical examination data on an annual basis. We estimate 75% of our study population will be African American, and 25% will be Caucasian. Both male and female populations will be represented, with the majority being female, as 50% of our patients will consist of those persons with gestational diabetes. We understand the final protocol will be based on the decision arrived at by the Steering Committee, and may involve the use of insulin-resistance-modifying drugs.